Estrogen Improves Insulin Sensitivity and Suppresses Gluconeogenesis Via the Transcription Factor Foxo1

Diabetes Volume 68, February 2019 291

Estrogen Improves Insulin Sensitivity and Suppresses Gluconeogenesis via the Transcription Factor Foxo1

Hui Yan,1 Wangbao Yang,1 Fenghua Zhou,1 Xiaopeng Li,1 Quan Pan,1 Zheng Shen,1 Guichun Han,2 Annie Newell-Fugate,2 Yanan Tian,2 Ravikumar Majeti,3 Wenshe Liu,4 Yong Xu,5 Chaodong Wu,1 Kimberly Allred,1 Clinton Allred,1 Yuxiang Sun,1 and Shaodong Guo1

Diabetes 2019;68:291–304 | https://doi.org/10.2337/db18-0638

Premenopausal women exhibit enhanced insulin sensi- be a potential approach to modulate glucose metabolism tivity and reduced incidence of type 2 diabetes (T2D) and prevent diabetes. compared with age-matched men, but this advantage disappears after menopause with disrupted glucose homeostasis, in part owing to a reduction in circulating The prevalence of type 2 diabetes (T2D) has shown sex b 17 -estradiol (E2). Fasting hyperglycemia is a hallmark of disparities, with a reduced incidence in women (1). Both T2D derived largely from dysregulation of hepatic glu- clinical and animal studies show a strong correlation be- cose production (HGP), in which Foxo1 plays a central tween estrogen deficiency and metabolic dysfunction (2,3). METABOLISM role in the regulation of gluconeogenesis. Here, we in- The reduction of estrogen in postmenopausal women vestigated the action of E2 on glucose homeostasis in accelerates the development of insulin resistance and male and ovariectomized (OVX) female control and liver- T2D (4). Clinical trials of estrogen replacement therapy fi speci c Foxo1 knockout (L-F1KO) mice and sought to in postmenopausal women demonstrated an amelioration understand the mechanism by which E2 regulates glu- of insulin resistance and reductions of plasma glucose level coneogenesis via an interaction with hepatic Foxo1. In and incidence of T2D (5,6). However, the potential risk of both male and OVX female control mice, subcutaneous breast cancer and stroke upon estrogen therapy under- E implant improved insulin sensitivity and suppressed 2 scores 17b-estradiol (E ) as a therapeutic agent (7,8). gluconeogenesis; however, these effects of E were 2 2 Thus, understanding of tissue-specificE action and its abolished in L-F1KO mice of both sexes. In our use of 2 molecular mechanism in metabolic regulation is required mouse primary hepatocytes, E2 suppressed HGP and gluconeogenesis in hepatocytes from control mice but for the development of targeted estrogen mimics convey- fi failed in hepatocytes from L-F1KO mice, suggesting that ing metabolic bene t without side effects. The maintenance of glucose homeostasis depends on Foxo1 is required for E2 action on the suppression of glucose uptake in muscle and adipose tissue and glucose gluconeogenesis. We further demonstrated that E2 suppresses hepatic gluconeogenesis through activation of production through gluconeogenesis and glycogenolysis in estrogen receptor (ER)a–phosphoinositide 3-kinase– liver (9). Estrogen reduces glucose level, but it is thought Akt–Foxo1 signaling, which can be independent of insulin to be derived from the promotion of glucose uptake in receptor substrates 1 and 2 (Irs1 and Irs2), revealing an muscle (10). However, some studies have also indicated important mechanism for E2 in the regulation of glucose that estrogen suppresses glucose production in the liver homeostasis. These results may help explain why pre- (11–13) and that impairing the estrogen receptor (ER)- menopausal women have lower incidence of T2D than mediated signaling results in hepatic insulin resistance and age-matched men and suggest that targeting ERa can hyperglycemia with elevated hepatic glucose production

1Department of Nutrition and Food Science, College of Agriculture and Life Corresponding author: Shaodong Guo, [email protected] Sciences, Texas A&M University, College Station, TX Received 14 June 2018 and accepted 3 November 2018 2Department of Physiology and Pharmacology, College of Veterinary Medicine and © 2018 by the American Diabetes Association. Readers may use this article as Biomedical Sciences, Texas A&M University, College Station, TX long as the work is properly cited, the use is educational and not for proﬁt, and the 3Department of Pharmaceutical Sciences, College of Pharmacy, Texas A&M work is not altered. More information is available at http://www.diabetesjournals University, College Station, TX .org/content/license. 4Department of Chemistry, Texas A&M University, College Station, TX 5Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX 292 Estrogen, Foxo1, and Glucose Homeostasis Diabetes Volume 68, February 2019

(HGP) (14,15). These lines of evidence underscore the im- determination of serum insulin, glucagon, and other hor- portance of estrogen signaling in the liver in control of mone levels by Bio-Plex mouse diabetes immunoassay glucose homeostasis; yet, the liver-specific action of estro- (Bio-Rad, Hercules, CA) and serum E2 concentration by gen on HGP and its mechanism remain unclear. an ELISA kit (Cayman, Ann Arbor, MI). Liver glycogen Foxo1, a member of O-class forkhead transcription content was measured from mice fasted 16 h as previously factor, is a predominant regulator for HGP, promoting described (24). gene transcription of the rate-limiting gluconeogenic en- zyme glucose-6-phosphatase (G6pc) (16,17). Foxo1 is sup- Cell Cultures and HGP Assay pressed by insulin through the activation of Akt via the Primary hepatocytes were isolated from 8- to 16-week-old insulin receptor substrate (Irs)1 and Irs2-associated phos- mice and cultured in DMEM with 10% FBS as previously phoinositide 3-kinase (PI3K) (18,19). In a T2D mouse described (21). Cells were serum starved for 6 h before model, insulin fails to activate Akt and suppress Foxo1, treatment with 100 nmol/L E2 and a variety of signal trans- resulting in enhanced HGP and hyperglycemia; this can duction inhibitors, which were applied 30 min prior to E2.For be prevented by deletion of hepatic Foxo1 (20). In this HGP assay, freshly isolated hepatocytes were cultured in study, we investigated the effect of E2 on the maintenance DMEM with 2% FBS. After 3 h of attachment, cells was of glucose homeostasis and the involvement of hepatic cultured in HGP buffer (120 mmol/L NaCl, 5.0 mmol/L KCl, Foxo1 in mice of both sexes. We further examined E2 2.0 mmol/L CaCl2, 25 mmol/L NaHCO3, 2.5 mmol/L regulatory mechanisms for HGP and gluconeogenesis in KH2PO4, 2.5 mmol/L MgSO4,10mmol/LHEPES,0.5% primary hepatocytes. In particular, we tested the hypoth- BSA, 10 mmol/L sodium DL-lactate, and 5 mmol/L pyruvate, esis that hepatic Foxo1 plays important roles in the E2 pH 7.4) and treated with chemicals. Culture medium was action of suppressing hepatic gluconeogenesis through collected to determine glucose production using an Amplex ER-associated Akt signaling. Red Glucose Assay kit (Invitrogen, Carlsbad, CA). Total HGP, glycogenolysis, and gluconeogenesis were determined and calculated as previously described (25). RESEARCH DESIGN AND METHODS Animals Quantitative Real-time PCR Animal protocols were approved by the institutional an- RNA was extracted with Trizol reagent (Invitrogen) and imal care and use committee at Texas A&M University. used for cDNA synthesis via an iScript cDNA Synthesis kit Liver-specific Foxo1 knockout (L-F1KO) mice were gener- (Bio-Rad). Gene expression was measured with SYBR L/L ated by breeding floxed Foxo1 mice with albumin-Cre Green Supermix in real-time PCR (Bio-Rad) using primers mice as previously described (21). The Cre recombinase as previously described and cyclophilin gene as an internal induced by the mouse albumin enhancer/promoter shows control (21). no influence on animal performance, as previously reported, and either the floxed-Foxo1L/L or albumin-Cre littermates Protein Immunoblotting were used as control mice in our studies (22,23). Eight- to An equal amount of protein was resolved in SDS-PAGE and 2 12-week-old control (Foxo1L/L) and L-F1KO (albumin-Cre+/ :: transferred to nitrocellulose membrane for Western blot. Foxo1L/L) mice on a mixed genetic background of C57BL/6J Antibodies for Foxo1, phosphorylated Foxo1 at Ser253, and 129/Sv were fed a standard chow diet (54% calories Akt, phosphorylated Akt at Ser473, Pck1, and GAPDH from carbohydrate, 14% from fat, and 32% from protein were purchased from Cell Signaling Technology. Protein and 3.0 kcal/g) (Envigo Teklad Diet) ad libitum. densitometry was performed and analyzed using ImageJ as previously described (21).

Ovariectomy and E2 Replacement Control and L-F1KO mice were randomly assigned to Statistical Analyses experimental groups (n =4–7). All mice were subjected Data were analyzed by one-way or two-way ANOVA to ﬁ to the subcutaneous implantation of placebo or E2 pellet determine the signi cance of the model as appropriate. (0.05 mg/pellet, 60-day release) (Innovative Research of The interaction between E2 and L-F1KO was analyzed in America, Sarasota, FL). Female mice underwent a bilateral two-way ANOVA to determine the dependency. Differ- ovariectomy (OVX) surgery (except for intact control mice) ences between groups were determined by Tukey post t P , at the time of pellet implantation. hoc test or Student test as appropriate. 0.05 was considered indicative of statistical signiﬁcance. Results are 6 Metabolic Analyses presented as means SEM. Blood glucose levels were measured using a glucometer (Bayer, Whippany, NJ). Glucose tolerance tests (GTTs) RESULTS and pyruvate tolerance tests (PTTs) were performed in E2 Regulates Glucose Homeostasis Depending on mice fasted for 16 h. Insulin tolerance tests (ITTs) were Hepatic Foxo1 in Male Mice conductedinmicefastedfor5h.Serumsampleswere We determined the role of estrogen in control of glu- collected from mice fasted for 16 h by cardiac puncture for cose homeostasis and its dependence on hepatic Foxo1. diabetes.diabetesjournals.org Yan and Associates 293

Generally, intact female mice exhibited a 16% lower fasting glucose compared with intact females (intact 51 6 glucose level compared with male mice (intact female 51 6 2.8 mg/dL vs. OVX 62.4 6 2.2 mg/dL, P , 0.01) (Fig. 6 P A A 2.8 mg/dL vs. male: 61 2.3 mg/dL, = 0.03) (Fig. 1 and 1 ). E2 reduced glucose levels by 21% in OVX control C). In female mice, OVX mice had a 22% increase in fasting females (OVX + placebo 62.4 6 2.2 mg/dL vs. OVX + E2

Figure 1—Estrogen regulates glucose homeostasis in mice. A–C: Blood glucose level in control and L-F1KO mice of both sexes fasted for 16 h overnight. Fasting glucose was measured every week, and the data show the most representative measurement. D–F: Glycogen content in the liver of mice fasted for 16 h overnight. Glycogen content was normalized by weight of liver. G–I: Body weight at 8 weeks after E2 implantation in control and L-F1KO mice of both sexes fasted overnight. J–L: Serum E2 level at 8 weeks after E2 implantation in mice fasted overnight. All data are expressed as the mean 6 SEM. For females, n =4–7, *P , 0.05 vs. OVX control mice with placebo. For males, n =4–5, *P , 0.05 vs. control mice with placebo. NS, not signiﬁcant (P . 0.05). *P , 0.05; **P , 0.01; ***P , 0.001. 294 Estrogen, Foxo1, and Glucose Homeostasis Diabetes Volume 68, February 2019

49.4 6 1.2 mg/dL, P , 0.001) and almost restored glucose mice displayed a decrease in gluconeogenesis in both sexes A B P , levels to those of intact mice (Fig. 1 and ). In control ( 0.05); however, the effect of E2 was abolished in H–L male mice, E2 decreased glucose levels by 16% (placebo L-F1KO mice (Fig. 2 ). OVX impaired insulin sensitivity 61.0 6 2.3 mg/dL vs. E2 51.0 6 2.0 mg/dL, P = 0.03) (Fig. as determined by ITT (Fig. 2M and P). E2 and L-F1KO 1C). Deﬁciency of hepatic Foxo1 decreased fasting glucose improved ITT in both sexes but had no additive effect (Fig. by 19% in OVX females and by 23% in males (Fig. 1B and 2N–R). Taken together, E2 improves insulin sensitivity and C). In L-F1KO mice, E2 led to a further 17% reduction of suppresses gluconeogenesis in a Foxo1-dependent manner. glucose levels in OVX females but failed to reduce glucose We measured mRNA levels of genes responsible for levels in males (Fig. 1B and C). These results indicate that insulin signaling and gluconeogenesis. The OVX surgery Irs2 hepatic Foxo1 is required for E2 to regulate glucose ho- suppressed hepatic mRNA expression by 40% and meostasis in males rather than females. induced G6pc and phosphoenolpyruvate carboxykinase To explain the sex disparity in glucose levels of L-F1KO 1(Pck1) expressions by 22% and 42%, respectively, in mice in response to E2, we measured fasting glycogen control females (P , 0.05); however, E2 induced Irs2 by content in the liver. Intact female mice showed a sevenfold 35% and suppressed G6pc and Pck1 by 22% and 30% in higher liver glycogen than that of males, indicating a higher OVX control females (Fig. 3A and B). E2 also induced Irs2 capacity of females to store glycogen and/or prevent by 50% and suppressed G6pc by 24% in male mice. The glycogenolysis in the liver than males (Fig. 1D and F). liver of L-F1KO mice displayed a reduction of Irs1 expres- In control females, OVX reduced liver glycogen by 80% sion by 45% in OVX females and 20% in males indepen- P , P , B C ( 0.01), but this effect of OVX was reversed by E2 ( dent of E2 treatment (Fig. 3 and ). Hepatic expression of 0.01) (Fig. 1D and E). L-F1KO increased liver glycogen by G6pc was reduced by 46–48% in both sexes upon deletion 2.2-fold in OVX females, and E2 further increased liver of liver Foxo1 (P , 0.05) (Fig. 3B and C). However, E2 glycogen by 2.5-fold in OVX L-F1KO females (Fig. 1E); failed to have the inhibitory effect on G6pc and Pck1 however, the effect of L-F1KO is masked by E2 when E2 is expressions in both sexes of L-F1KO mice (Fig. 3B and C). abundantly available. Collectively, the sex differences in Levels of metabolic hormones were determined in the liver glycogen may help explain the sex disparity in glucose serum from mice fasted for 16 h. OVX mice exhibited in- homeostasis in response to E2. creases in leptin and plasminogen activator inhibitor 1 levels Compared with intact female mice, OVX increased by 3.0- and 1.3-fold, respectively, compared with intact bodyweightby40%(Fig.1G), while E2 reduced body female mice, while E2 implant decreased levels of leptin weight of OVX females by 14% in both control and andplasminogenactivatorinhibitor1by56%and30%in L-F1KO mice (Fig. 1H). Body weight was unaffected by OVX mice (P , 0.05) (Fig. 3D and E). In male mice, deletion E2 in male mice (Fig. 1I). Deletion of hepatic Foxo1 had of liver Foxo1 increased glucagon and ghrelin levels by no effect on body weight in both sexes (Fig. 1G–I). OVX 5.1- and 2.6-fold, respectively (P , 0.05) (Fig. 3F). E2 mice exhibited a 44% lower serum E2 level compared with increased ghrelin level by 1.8-fold in control males (P , intact females owing to absence of ovarian estrogen 0.05) (Fig. 3F). Although sex differences exist in the re- J ﬁ production (Fig. 1 ). E2 implant signi cantly increased lease of metabolic hormones, the increased glucagon and serum E2 levels of control and L-F1KO mice in both sexes ghrelin levels, as well as decreased leptin level, may be (P , 0.05) (Fig. 1K and L). Interestingly, in males, serum a result of lower blood glucose upon E2 stimulation and E2 concentration was undetectable in control mice, but ablation of Foxo1, which suggests a feedback regulation in L-F1KO males exhibited a detectable serum E2 value at glucose homeostasis. 1.56 6 0.53 pg/mL (Fig. 1L). Hepatic Foxo1 Is Required for Estrogen to Suppress Hepatic Foxo1 Is Necessary for Estrogen to Improve HGP and Gluconeogenesis in Primary Hepatocytes Insulin Sensitivity and Suppress Gluconeogenesis in We next assessed E2 action on hepatic glucose metabolism Mice in primary hepatocytes. L-F1KO hepatocytes exhibited We next examined the role of hepatic Foxo1 in E2 action on a 32% lower HGP after 1 h culture in assay buffer com- insulin sensitivity and gluconeogenesis in mice. Surgical pared with control cells (P , 0.05) (Fig. 4A). E2 decreased removal of the ovaries impaired glucose clearance after HGP by 14% and 20% after 3- and 6-h culture in control intraperitoneal injection of glucose (Fig. 2A and D). Either cells, respectively (P , 0.05), but failed to suppress HGP in E2 implant or deletion of liver Foxo1 resulted in an L-F1KO cells (Fig. 4A). We further divided HGP into improved glucose tolerance in both sexes, but there gluconeogenesis and glycogenolysis in response to E2. were no additive effects of these two in L-F1KO mice Although E2 decreased glycogenolysis by 25% in control (Fig. 2B–F). We examined whether estrogen-lowered fast- hepatocytes (P , 0.05), the contribution of glycogen ing glucose is derived from suppression of gluconeogene- breakdown to total HGP was much smaller than that of sis, as determined by PTT. OVX had impaired pyruvate gluconeogenesis (Fig. 4B). Thus, 18% reduction of gluco- tolerance compared with control intact females (P , 0.05) neogenesis by E2 mainly contributed to suppression of (Fig. 2G and J), while E2 suppressed gluconeogenesis in HGP in control cells (P , 0.05) (Fig. 4B). Consistently, E2 both sexes of control mice (P , 0.05) (Fig. 2H–L). L-F1KO reduced expressions of G6pc by 25% and Pck1 by 30% in diabetes.diabetesjournals.org Yan and Associates 295

Figure 2—Estrogen improves insulin sensitivity and suppresses gluconeogenesis in mice. A–C: GTT was performed in control and L-F1KO mice of both sexes fasted for 16 h overnight. Glucose was administered at 2 g/kg body wt of mice by intraperitoneal injection after 16 h overnight fasting, and glucose level was measured at indicated time points. D–F: Area under the curve (AUC) of the GTT displayed in A–C, respectively. G–I: PTT was performed on mice fasted for 16 hours overnight. Pyruvate was administered at 2 g/kg body wt by intraperitoneal injection after overnight fasting, and glucose level was measured at indicated time points. J–L: Area under the curve of the PTT showed in G–I, respectively. M–O: ITT was performed on mice fasted for 5 h. Insulin was administered at 2 units/kg body wt by intraperitoneal injection, and glucose level was measured at indicated time points. P–R: Area under the curve of the ITT shown in M–O, respectively. All data are expressed as mean 6 SEM. For females, n =4–7, *P , 0.05 vs. OVX control mice with placebo. n =4–7, #P , 0.05 vs. intact mice. For males, n =4–5, *P , 0.05 vs. control mice with placebo. NS, not signiﬁcant (P . 0.05). *P , 0.05; **P , 0.01; ***P , 0.001. 296 Estrogen, Foxo1, and Glucose Homeostasis Diabetes Volume 68, February 2019

Figure 3—Estrogen suppresses expression of gluconeogenic genes in liver and inﬂuences serum hormone levels associated with glucose metabolism in mice. A–C: Relative mRNA expressions in the liver of control and L-F1KO mice of both sexes fasted for 16 hours overnight after 8 weeks of pellet implantation. D–F: Serum hormone levels in mice fasted for 16 h overnight. Blood samples were collected by cardiac puncture in euthanized mice. All data are expressed as the mean 6 SEM. For females, n =4–7, *P , 0.05 vs. OVX control mice on placebo. For males, n =4–5, *P , 0.05 vs. control mice on placebo. *P , 0.05; **P , 0.01; ***P , 0.001.

control hepatocytes (P , 0.05) (Fig. 4C). L-F1KO hepato- Akt phosphorylates Foxo1 at Ser253,thuspromoting cytes showed reductions of G6pc and Pck1 expressions by Foxo1 nuclear export and/or degradation (26). E2 in- 55% compared with control cells (P , 0.05); however, E2 creased phosphorylation of Akt and Foxo1 by 1.7- and failed to further suppress these genes’ expression in 1.5-fold, respectively. The total Foxo1 degradation was L-F1KO cells (Fig. 4C). unnoticeable by E2 treatment alone but evident by E2 diabetes.diabetesjournals.org Yan and Associates 297

Figure 4—Estrogen suppresses HGP and gluconeogenesis depending on hepatic Foxo1 in primary hepatocytes. A: HGP in primary hepatocytes from control and L-F1KO mice. HGP was measured at indicated time points after 0.1 mmol/L E2 stimulation, and normalized to total protein levels. B: Glucose production in primary hepatocytes from control and L-F1KO mice upon 0.1 mmol/L E2 stimulation in the presence (HGP) or absence (glycogenolysis) of pyruvate and lactate. The difference between these two values was interpreted to represent gluconeogenesis. C: Relative mRNA levels of gluconeogenic genes in primary hepatocytes from control and L-F1KO mice upon 0.1 mmol/L E2 stimulation for 3 h. D and E: Western blots (D) and corresponding quantiﬁcation (E) of insulin signaling protein in hepatocytes from control mice upon 1 h stimulation of 0.1 mmol/L E2 or 0.1 mmol/L insulin. p-, phosphorylated. F: Glucose production in primary hepatocytes from control and Foxo1 S253A knock-in mice (S253A) upon 0.1 mmol/L E2 stimulation in the presence (HGP) or absence (glycogenolysis) of pyruvate and lactate (glycogenolysis). All data are expressed as the mean 6 SEM. NS, not signiﬁcant (P . 0.05). *P , 0.05, **P , 0.01, ***P , 0.001 vs. control + vehicle. n = 3 of each group.

and insulin cotreatments (Fig. 4D and E). To conﬁrm the E2 Suppresses Gluconeogenesis Dependent on PI3K- importance of phosphorylation of Foxo1 at Ser253 for Akt Signaling but Independent of Irs1 and Irs2 253 E2 action, we generated Foxo1 Ser knock-in mice We next determined whether activation of Akt signaling (S253A), in which serine 253 was replaced with alanine downstream from PI3K is necessary for E2 in suppression to prevent Akt phosphorylation (K. Zhang, X. Guo, of Foxo1 and gluconeogenesis in hepatocytes using Akt1/2 H. Yan, Y. Wu, , Q. Pan, Z. Shen, X. Li, Y. Chen, L. Li, kinase inhibitor (Akti) or PI3K inhibitor wortmannin Y. Qi, Z. Xu, W. Xie, W. Zhang, D. Threadgill, L. He, Y. Sun, (Wort). Both Akti and Wort completely prevented E2 ef- M. F. White, H. Zheng, and S. Guo, unpublished obser- fects on HGP, glycogenolysis, and gluconeogenesis (Fig. A G6pc Pck1 vations). Hepatocytes from S253A mice exhibited in- 5 ), as well as E2-suppressed and mRNA ex- creases in HGP by 31% and gluconeogenesis by 36% pression and Pck1 protein levels (Fig. 5B and C). Mean- owing to the blockage of Foxo1 phosphorylation at while, E2-induced reduction of Pck1 protein level and 253 Ser (P , 0.05); meanwhile, E2 failed to reduce phosphorylation of both Akt and Foxo1 were also blocked HGP in S253A hepatocytes (Fig. 4F). These data sug- by these two inhibitors (Fig. 5C and D). Additionally, Akti gested that phosphorylation of Foxo1 Ser253 by Akt is alone showed increases in basal levels of G6pc and Pck1 but ﬁ required for E2 in the regulation of Foxo1 activity and no signi cant effect on HGP or gluconeogenesis in the hepatic gluconeogenesis. primary hepatocytes (Fig. 5A–D). 298 Estrogen, Foxo1, and Glucose Homeostasis Diabetes Volume 68, February 2019

Figure 5—Activation of Akt signaling is required for estrogen to suppress gluconeogenesis. A: Glucose release in primary hepatocytes from control mice in the presence (HGP) or absence (glycogenolysis) of pyruvate and lactate. Cells were treated with 10 mmol/L Akti or 0.2 mmol/L Wort 30 min prior to 0.1 mmol/L E2 stimulation. B: Relative mRNA levels of gluconeogenic genes in primary hepatocytes from control mice upon stimulation of 10 mmol/L Akti and 0.2 mmol/L Wort 30 min prior to 0.1 mmol/L E2. C and D: Western blots (C) and corresponding quantiﬁcation (D) of insulin-signaling protein in hepatocytes from control mice upon stimulation of 10 mmol/L Akti and 0.2 mmol/L Wort 30 min prior to 0.1 mmol/L E2. p-, phosphorylated. E: Blood glucose levels of control and L-DKO mice in random-fed or fasted state upon OVX surgery and E2 implant. Blood glucose levels were measured in littermates of male, intact female, and OVX female mice at 12 weeks of age random fed or after 16 h overnight fast. F: Glucose production in primary hepatocytes from control and L-DKO mice upon 0.1 mmol/L E2 or 0.1 mmol/L ERa agonist PPT stimulation in the presence (HGP) or absence (glycogenolysis) of pyruvate and lactate. All data are expressed as the mean 6 SEM. NS, not signiﬁcant (P . 0.05). *P , 0.05, **P , 0.01, ***P , 0.001 vs. control + vehicle. n = 3 of each group. diabetes.diabetesjournals.org Yan and Associates 299

Moreover, we further determined whether Irs1 and Irs2 PPT enhanced phosphorylation levels of Akt and Foxo1 P , are required for E2-mediated glucose homeostasis using by 2.7-fold and 1.8-fold, respectively ( 0.05) and liver-specific Irs1 and Irs2 double knockout (L-DKO) mice suppressed expression of G6pc by 30% in control hepato- as we previously characterized (27). In a random-fed state, cytes (P , 0.05) (Fig. 6B and C). L-DKO male mice displayed hyperglycemia with 158% in- We finally determined whether overexpression of ERa creased glucose level (control 124.7 6 4.38 mg/dL vs. L-DKO enhances insulin- or E2-induced Akt and Foxo1 phosphor- 321.50 6 51.73 mg/dL, P = 0.01); however, L-DKO females ylation in HepG2 cells, human hepatoma cells. Insulin and 6 exhibited euglycemia (control 119.9 4.37 mg/dL vs. L-DKO E2 increased Akt phosphorylation by 10-fold and 5.4-fold, 135.0 6 6.93 mg/dL, P =0.2)(Fig.5E). In the fasting state, respectively, and induced Foxo1 phosphorylation by 2.9- L-DKO males had increased glucose level by 62% (control fold and 2.8-fold (P , 0.05). Overexpression of ERa 54.2 6 1.9 mg/dL vs. L-DKO 87.6 6 5.5 mg/dL, P , 0.01), enhanced basal phosphorylation of Akt and Foxo1 by while L-DKO female mice only had increased glucose level by 3.4-fold and 2.0-fold. Strikingly, ERa overexpression fur- 6 6 29% (control 44.8 1.65 mg/dL vs. L-DKO 57.8 ther enhanced insulin- and E -stimulated Akt phosphor- P , E fi 2 2.63 mg/dL, 0.01) (Fig. 5 ). Signi cantly, L-DKO females ylation by 67% and 42% and Foxo1 phosphorylation by developed hyperglycemia 2 weeks after OVX, with 56% in- 1.3- and 1.2-fold (P , 0.05) (Fig. 6D). creased glucose, in the random-fed state (intact L-DKO Taken together, these results indicate that ERa is nec- 6 6 135.0 6.9 mg/dL vs. OVX L-DKO 210.3 13.5 mg/dL, essary and sufficient for E to activate Akt-Foxo1 signaling P , 2 0.01), while OVX only increased the glucose level by and suppress gluconeogenesis. 21% in control females (intact control 119.9 6 4.4 mg/dL vs. OVX control 136.2 6 5.9 mg/dL, P = 0.04) (Fig. 5E). Collectively, for OVX females, random-fed L-DKO mice had DISCUSSION a 54% increase in glucose level (control 136.2 6 5.91 mg/dL Estrogen has been shown to be involved in both cen- vs. L-DKO 210.3 6 13.5 mg/dL, P = 0.04) and a 43% tral and peripheral regulation in glucose homeostasis increase in fasting glucose (control 60.4 6 5.36 mg/dL vs. (29,30). Estrogen deficiency or impaired estrogen signaling L-DKO 86.0 6 4.89 mg/dL, P = 0.03) compared with control is associated with insulin resistance and dysregulation of E mice (Fig. 5 ). Moreover, subcutaneous implantation of E2 metabolic homeostasis, thus contributing to the develop- significantly decreased blood glucose of both control and ment of T2D and obesity in both human and animal models L-DKO OVX mice in both fed and fasting states and totally (31–33). However, the contribution of tissue-specific action E abolished the effect of L-DKO on glucose level (Fig. 5 ). of estrogen to metabolic changes and underlying mecha- Additionally, in in vitro HGP assay, L-DKO primary hep- nisms have yet to be elucidated. In this study, we have atocytes exhibited increases in HGP, glycogenolysis, and established that activation of ERa-Akt-Foxo1 signaling is an P , gluconeogenesis by 65%, 27%, and 71%, respectively ( important mechanism for estrogen in the liver in main- 0.05). Moreover, E2 inhibited HGP and gluconeogenesis by taining glucose homeostasis. F 17% and 20%, respectively, in L-DKO cells (Fig. 5 ). These In general, glucose homeostasis is maintained by glucose results suggested that ovarian hormone protects against uptake in muscle and adipose tissue and endogenous glu- hyperglycemia induced by loss of hepatic insulin function cose production in the liver. Previous studies reported that and that E2-mediated suppression of hepatic gluconeogen- estrogen lowers glucose level, which is associated with esis requires the activation of PI3K-Akt-Foxo1 signaling but enhanced glucose uptake in muscle through activation of in an Irs1- and Irs2-independent manner. Akt and induction of GLUT4 expression (10,34). However, blockage of E signaling in global or liver ERa knockout a fi 2 Activation of ER Is Necessary and Suf cient to mice causes hepatic insulin resistance and enhanced HGP, Activate Akt-Foxo1 Signaling and Suppress resulting in hyperglycemia (14,35). In this study, E im- Gluconeogenesis in Hepatocytes 2 ERa is the major form of ERs in the liver (28). We next proved glucose tolerance in both sexes; however, such an effect was blocked upon deletion of hepatic Foxo1. Our investigated whether ERa modulates E2 effect on suppression of HGP in hepatocytes using its specific agonist and results supported that improvement of glucose homeostasis antagonist. ERa-selective antagonist methylpiperidinopyr- by E2 is regulated by hepatic Foxo1-mediated gluconeogenesis rather than by promoting muscle glucose uptake (12). azole (MPP) completely blocked E2-reduced HGP, gluconeogenesis (Fig. 6A), and expression of G6pc and Pck1 (Fig. In liver, Foxo1 promotes gluconeogenesis through activat- G6pc Pck1 6B), as well as phosphorylation of Akt and Foxo1 (Fig. 6C). ing transcription of and by directly binding to Moreover, MPP alone reduced basal phosphorylation of insulin response element CAAAACAA on their promoter Akt and Foxo1 by 30% and 55%, respectively (P , 0.05) regions (26,36). Liver-specific ablation of Foxo1 impairs (Fig. 6C); this resulted in increases in G6pc and Pck1 gluconeogenesis and reduces fasting glucose level (37). De- levels by 68% and 25% (P , 0.05) (Fig. 6B). ERa agonist letion of hepatic Foxo1 protects diabetic mice, such as propylpyrazoletriol (PPT) showed an effect similar to that L-DKO mice or db/db mice, from hepatic insulin resistance of E2 to reduce HGP and gluconeogenesis in both control (23,27). In this study, suppression of gluconeogenesis by E2 (Fig. 6A) and L-DKO (Fig. 5E) hepatocytes. Meanwhile, was abolished upon deletion of Foxo1; this supported 300 Estrogen, Foxo1, and Glucose Homeostasis Diabetes Volume 68, February 2019

Figure 6—ERa is required and sufﬁcient to activate Akt-Foxo1 signaling and suppress gluconeogenesis in primary hepatocytes. A: Glucose production in primary hepatocytes from control mice in the presence (HGP) or absence (glycogenolysis) of pyruvate and lactate. Cells were treated with 1 mmol/L ERa antagonist MPP prior to stimulation of 0.1 mmol/L E2 or 0.1 mmol/L ERa agonist PPT. B: Relative mRNA levels of diabetes.diabetesjournals.org Yan and Associates 301

a notion that hepatic Foxo1 is required for E2 action in the action; this suggested that the interaction of ERa with modulation of gluconeogenesis. PI3K may contribute to activation of Akt-Foxo1 and This study illustrated that E2 inhibits Foxo1 trans- regulation of glucose homeostasis independent of Irs1 activation via the interaction with ERa and activation of and Irs2. The phosphorylation sites of Foxo1 by Akt are PI3K-Akt signaling, thus serving as a molecular basis for also identified in other Foxo members, including Foxo3 a and Foxo4. Our previous study indicated that Foxo1, E2 action on suppression of hepatic gluconeogenesis. ER is involved in glucose homeostasis, and impaired ERa rather than Foxo3 and -4, significantly influences glucose fi function is associated with insulin resistance, T2D, and homeostasis(21).EvenifE2 modi es the activity of other metabolic syndrome in both human and animal models Foxo isoforms via Akt, the contribution of Foxo3 and (14,15,35). Our results demonstrated that activation Foxo4 by E2 to glucose metabolism is limited. E exhibits a Foxo1-independent mechanism to reduce of ERa is required for E2 to suppress HGP and that 2 glucose level only in females. OVX decreased hepatic overexpression of ERa enhances insulin- and E2-induced phosphorylation of Foxo1. Ambiguity about how ERa glycogen (46), which was restored by E2, as previously modulates insulin signaling and glucose metabolism still reported (11,12,47,48). E2 increased glycogen independent remains. A recent study indicated that an estrogen re- of hepatic Foxo1, which may account for the reduced sponse element half-site (AGGTCA) presents in promoter glucose level in L-F1KO females. Hepatic glycogen is regions of G6pc and Pck1 (38). Recruitment of ERa to regulated by glycogen synthase kinase 3a (GSK3a), which those promoters upon E activation inhibits transcrip- phosphorylates glycogen synthase and inhibits glycogen 2 a tions of G6pc and Pck1 (35), but Yasrebi et al. (39) synthesis (49). GSK3 is inactivated by Akt-mediated phosphorylation, and loss of function of GSK3a is asso- reported that E2 still decreased fasting glucose and improved glucose tolerance in the ERa knock-in mice, in ciated with increased hepatic glycogen and lowered glucose which the knock-in mutation blocks the functional es- levels (50). Thus, we speculate that activation of Akt and a trogen response element binding activity for DNA, ruling inhibition of GSK3 to promote glycogen synthesis by E2 may be involved in a Foxo1-independent mechanism for E out this genomic signaling of ERa.AnE2-dependent 2 binding of ERa with Foxo1 was reported to suppress to reduce glucose level in females. In fact, a study reported Foxo1 transactivation in the MCF7 cell line (40). In- that hepatic Foxo1 ablation is associated with an increase hibition of Akt signaling by inhibitors or by S253A in liver glycogen and inhibition of glycogenolysis in hep- mutant of the Akt phosphorylation site in Foxo1 totally atocytes (37); another study indicated that Foxo1 has blocked effects of estrogen on HGP; this suggested the limited effects on hepatic glycogen (20). Nevertheless, 253 we found that deficiency of hepatic Foxo1 increased liver importance of indirect E2 action on Foxo1 Ser phosphorylation via ERa-PI3K-Akt rather than a direct in- glycogen only in OVX females, which is masked by E2 teraction between ERa and Foxo1. Direct interactions of implant. That deletion of hepatic Foxo1 increased liver ERa with several proteins in the insulin signaling cascade glycogen might be the result of inhibition of gluconeogen- have been reported. ERa binds to Irs1 and Irs2 and esis and accumulation of glucose-6-phosphate, a substrate promotes stability of the ERa/Irs complex in breast for glycogen synthesis. cancer lines (41,42); it also binds to the p85a catalytic Sex differences exist in liver glycogen levels and subunit of PI3K in human embryonic kidney (HEK)293T, responses to E2. Male mice exhibited an 85% reduction MCF7, and aortic endothelial cells (43,44). A direct in- in liver glycogen compared with intact females, and E2- teraction between p85a and ERa in proopiomelanocortin increased hepatic glycogen levels as observed in OVX (POMC) progenitor neurons has been reported, and de- females disappeared in male mice. Several studies have letion of PI3K in POMC neurons attenuates E2-reduced indicated that long-term exercise training increases glyco- glucose level in OVX females (45). However, the question gen storage in both muscle and liver with enhanced glycogen of whether these protein interactions contribute to he- synthase activity (51). Female mice naturally exhibited patic glucose metabolism remains. In this study, L-DKO a higher level of physical activity than male mice (52); we intact females were protected from hyperglycemia and suspect that such a higher physical activity may contribute deletion of Irs1 and Irs2 did not disrupt E2-suppressed to a greater capacity to store glycogen. This also implies a fasting glucose level in vivo or hepatic gluconeogenesis sexual dimorphism in the contribution of glycogen metab- in vitro, whereas PI3K inhibitor totally blocked the E2 olism to glucose homeostasis.

gluconeogenic genes in primary hepatocytes from control mice upon stimulation of 0.1 mmol/L E2,1mmol/L MPP, or 0.1 mmol/L PPT. C: Western blots and corresponding quantification of insulin-signaling protein in hepatocytes from control mice upon stimulation of 0.1 mmol/L E2,1mmol/L MPP, or 0.1 mmol/L PPT. D: Western blots and corresponding quantification of Foxo1 protein in HepG2 cells upon overexpression of ERa and stimulation of E2 or insulin. HepG2 cells were transfected with 10 mg plasmid DNA expressing green fluorescent protein (GFP) or ERa for 30 h, followed by 6 h starvation prior to 0.1 mmol/L E2 or 0.1 mmol/L insulin stimulation for 1 h. All data are expressed as mean 6 SEM. *P , 0.05, **P , 0.01, ***P , 0.001 vs. control + vehicle. ###P , 0.001 vs. control + ERa overexpression. n = 3 of each group. p- or p, phosphorylated. 302 Estrogen, Foxo1, and Glucose Homeostasis Diabetes Volume 68, February 2019

E2 suppressed body weight in OVX females independent from androstenedione by aromatization in the ovaries of hepatic Foxo1 but reduced leptin level dependent on andisreleasedascirculatinghormoneactingondistant Foxo1, whereas E2 was less effective in reduction of body tissues. In men, E2 is produced locally in extragonadal weight and leptin in males. Our previous studies showed tissues by aromatization from testosterone (31,56). that estrogen increases energy expenditure and decreases Patients lacking aromatase function and aromatase- body weight through binding to its receptors in the central deficient mice (ArKO mice), with endogenous estradiol nervous system (CNS) (53). In the OVX female mice, synthesis diminished in both cases, exhibit impaired obesity developed and E2 administration reduced the glucose homeostasis (3,57). Men lacking aromatase also body weight and serum leptin level, suggesting E2 increases exhibit hyperinsulinemia (58), and male ArKO mice de- the leptin sensitivity, which is likely via increasing expres- velop insulin resistance (57); this suggests that E2 sig- sion of Irs2, a modulator of leptin action in the CNS (54). naling is similarly crucial in metabolic homeostasis in Leptin level is highly associated with the amount of fat both sexes. Of note, L-F1KO males had a detectable serum mass. E2-decreased fat mass accumulation through CNS E2 level compared with the undetectable serum E2 in control may explain the decreased leptin level. However, E2 ad- males. Given that deletion of hepatic Foxo1 increases ministration reduced the body weight in L-F1KO mice; the cholesterol levels (59) and cholesterol is a precursor for effect can be independent of leptin level but dependent on steroid hormone biosynthesis, we speculate that hepatic leptin sensitivity with an increase of Irs2 expression and Foxo1 may also be involved in estrogen biosynthesis in- energy expenditure (54). The sex differences may result directly through the regulation of cholesterol production in from differences in expression of ER isoforms in various the liver. tissues (55), absorbance rates of exogenous E2 and local E2 In summary, our study demonstrated that E2 improves concentrations, and capacity of E2 to cross the blood-brain insulin sensitivity and suppresses hepatic gluconeogenesis barrier to CNS. through inhibition of Foxo1 via activation of ERa-PI3K- Estrogen is critical to maintaining metabolic homeo- Akt signaling (Fig. 7). PI3K is required for E2 action, but stasis in both sexes (31). In this study, E2 implant comprehensive studies into the direct interaction of ERa improved insulin sensitivity and lowered fasting glucose with PI3K regulating hepatic glucose metabolism have not in both sexes. In premenopausal women, E2 is converted been conducted. Since estrogen promotes the development of the reproductive system, the metabolism and reproductive biology can be regulated by different sets of target genes. Therefore, the identification of tissue-specific Insulin actions of E2 and direct targets of ERs will facilitate the IR development of novel selective ligands that prevent T2D, E2 cardiovascular disease, and obesity without promoting abnormal sex characteristics or breast cancer. IRS1/2 p85 pY p110 PI3K Acknowledgments. The authors thank Jennifer DeLuca (Department of Nutrition and Food Science, College of Agriculture and Life Sciences, Texas A&M University) for technical assistance for the OVX surgery. E2 PDK1 Funding. This work was supported by National Institutes of Health, National ERαER pT308 Institute of Diabetes and Digestive and Kidney Diseases, grants (RO1-DK-095118 Akt pS473 and R56-DK-118334-01), American Diabetes Association Career Development Award 1-15-CD-09, an American Heart Association grant (BGIA-7880040), Faculty Start-up funds from Texas A&M University Health Science Center and AgriLife Nucleus Research, and a U.S. Department of Agriculture National Institute of Food and pS253 Agriculture grant (Hatch 1010958) to S.G. S.G. is the recipient of the 2015 Amer- Foxo1 G6pc ican Diabetes Association Research Excellence Thomas R. Lee Award. Duality of Interest. No potential conflicts of interest relevant to this article CAAAACAACAAAACAA Pck1 were reported. Author Contributions. H.Y., W.Y., and F.Z. designed and conducted experiments and performed data analyses. H.Y. and S.G. wrote the manuscript. X.L., Q.P., Z.S., and K.A. conducted experiments. G.H., A.N.-F., Y.T., R.M., W.L., Figure 7—Schematic diagram represents the role of estrogen in the Y.X., C.W., C.A., and Y.S. reviewed and edited the manuscript. S.G. supervised the regulation of glucose metabolism. Estrogen suppresses hepatic project, conceived of the hypothesis, and designed experiments. S.G. is the gluconeogenesis and lowers blood glucose through interaction guarantor of this work and, as such, had full access to all the data in the study and a with ER in a Foxo1-dependent manner. E2 inhibits Foxo1 and its takes responsibility for the integrity of the data and the accuracy of the data target G6pc expression indirectly, depending on the activation of analysis. PI3K-Akt signaling. IR, insulin receptor; p, phosphorylated; PIP2, Prior Presentation. Parts of this study were presented in abstract form at phosphatidylinositol (4,5)-bisphosphate; PIP3, phosphatidylinositol (3,4,5)-trisphosphate; PDK1, 3-phosphoinositide-dependent protein the 78th Scientific Sessions of the American Diabetes Association, Orlando, FL, kinase 1; Y, tyrosin; T, threonine; S, serine. 22–26 June 2018. diabetes.diabetesjournals.org Yan and Associates 303

References 20. O-Sullivan I, Zhang W, Wasserman DH, et al. FoxO1 integrates direct and 1. Danaei G, Finucane MM, Lu Y, et al.; Global Burden of Metabolic Risk indirect effects of insulin on hepatic glucose production and glucose utilization Factors of Chronic Diseases Collaborating Group (Blood Glucose). National, [published correction appears in Nat Commun 2015;6:7861]. Nat Commun 2015; regional, and global trends in fasting plasma glucose and diabetes prevalence 6:7079 since 1980: systematic analysis of health examination surveys and epidemiological 21. Zhang K, Li L, Qi Y, et al. Hepatic suppression of Foxo1 and Foxo3 studies with 370 country-years and 2$7 million participants. Lancet 2011;378: causes hypoglycemia and hyperlipidemia in mice. Endocrinology 2012;153: 31–40 631–646 2. Salpeter SR, Walsh JM, Ormiston TM, Greyber E, Buckley NS, Salpeter EE. 22. Michael MD, Kulkarni RN, Postic C, et al. Loss of insulin signaling in Meta-analysis: effect of hormone-replacement therapy on components of the hepatocytes leads to severe insulin resistance and progressive hepatic dys- metabolic syndrome in postmenopausal women. Diabetes Obes Metab 2006;8: function. Mol Cell 2000;6:87–97 538–554 23. Dong XC, Copps KD, Guo S, et al. Inactivation of hepatic Foxo1 by insulin 3. Misso ML, Murata Y, Boon WC, Jones ME, Britt KL, Simpson ER. Cellular and signaling is required for adaptive nutrient homeostasis and endocrine growth molecular characterization of the adipose phenotype of the aromatase-deficient regulation. Cell Metab 2008;8:65–76 mouse. Endocrinology 2003;144:1474–1480 24. Lo S, Russell JC, Taylor AW. Determination of glycogen in small tissue 4. Louet J-F, LeMay C, Mauvais-Jarvis F. Antidiabetic actions of estrogen: samples. J Appl Physiol 1970;28:234–236 insight from human and genetic mouse models. Curr Atheroscler Rep 2004;6: 25. Pajvani UB, Shawber CJ, Samuel VT, et al. Inhibition of Notch signaling 180–185 ameliorates insulin resistance in a FoxO1-dependent manner. Nat Med 2011;17: 5. Kim C, Kong S, Laughlin GA, et al.; Diabetes Prevention Program Research 961–967 Group. Reductions in glucose among postmenopausal women who use and do not 26. Guo S, Rena G, Cichy S, He X, Cohen P, Unterman T. Phosphorylation of use estrogen therapy. Menopause 2013;20:393–400 serine 256 by protein kinase B disrupts transactivation by FKHR and mediates 6. Manson JE, Chlebowski RT, Stefanick ML, et al. The Women’sHealth effects of insulin on insulin-like growth factor-binding protein-1 promoter activity Initiative Hormone Therapy Trials: Update and Overview of Health Outcomes through a conserved insulin response sequence. J Biol Chem 1999;274:17184– During the Intervention and Post-Stopping Phases. JAMA 2013;310:1353– 17192 1368 27. Guo S, Copps KD, Dong X, et al. The Irs1 branch of the insulin signaling 7. Viscoli CM, Brass LM, Kernan WN, Sarrel PM, Suissa S, Horwitz RI. A clinical cascade plays a dominant role in hepatic nutrient homeostasis. Mol Cell Biol 2009; trial of estrogen-replacement therapy after ischemic stroke. N Engl J Med 2001; 29:5070–5083 345:1243–1249 28. Kuiper GG, Carlsson B, Grandien K, et al. Comparison of the ligand binding 8. Stefanick ML, Anderson GL, Margolis KL, et al.; WHI Investigators. Effects of specificity and transcript tissue distribution of estrogen receptors alpha and beta. conjugated equine estrogens on breast cancer and mammography screening in Endocrinology 1997;138:863–870 postmenopausal women with hysterectomy. JAMA 2006;295:1647–1657 29. Barros RP, Gustafsson JA. Estrogen receptors and the metabolic network. 9. Pilkis SJ, Granner DK. Molecular physiology of the regulation of hepatic Cell Metab 2011;14:289–299 gluconeogenesis and glycolysis. Annu Rev Physiol 1992;54:885–909 30. Faulds MH, Zhao C, Dahlman-Wright K, Gustafsson JA. The diversity of sex 10. Moreno M, Ordoñez P, Alonso A, Díaz F, Tolivia J, González C. Chronic 17beta- steroid action: regulation of metabolism by estrogen signaling. J Endocrinol 2012; estradiol treatment improves skeletal muscle insulin signaling pathway components 212:3–12 in insulin resistance associated with aging. Age (Dordr) 2010;32:1–13 31. Mauvais-Jarvis F. Estrogen and androgen receptors: regulators of fuel 11. Matute ML, Kalkhoff RK. Sex steroid influence on hepatic gluconeogenesis homeostasis and emerging targets for diabetes and obesity. Trends Endocrinol and glucogen formation. Endocrinology 1973;92:762–768 Metab 2011;22:24–33 12. Sladek CD. The effects of human chorionic somatomammotropin and es- 32. Hevener AL, Clegg DJ, Mauvais-Jarvis F. Impaired estrogen receptor action in tradiol on gluconeogenesis and hepatic glycogen formation in the rat. Horm Metab the pathogenesis of the metabolic syndrome. Mol Cell Endocrinol 2015;418:306– Res 1975;7:50–54 321 13. Carter S, McKenzie S, Mourtzakis M, Mahoney DJ, Tarnopolsky MA. Short- 33. Zhu L, Brown WC, Cai Q, et al. Estrogen treatment after ovariectomy protects term 17b-estradiol decreases glucose R(a) but not whole body metabolism during against fatty liver and may improve pathway-selective insulin resistance. Diabetes endurance exercise. J Appl Physiol (1985) 2001;90:139–146 2013;62:424–434 14. Bryzgalova G, Gao H, Ahren B, et al. Evidence that oestrogen receptor-alpha 34. Gorres BK, Bomhoff GL, Morris JK, Geiger PC. In vivo stimulation of oestrogen plays an important role in the regulation of glucose homeostasis in mice: insulin receptor a increases insulin-stimulated skeletal muscle glucose uptake. J Physiol sensitivity in the liver. Diabetologia 2006;49:588–597 2011;589:2041–2054 15. Gao H, Fält S, Sandelin A, Gustafsson JA, Dahlman-Wright K. Genome-wide 35. Qiu S, Vazquez JT, Boulger E, et al. Hepatic estrogen receptor a is critical for identification of estrogen receptor alpha-binding sites in mouse liver. Mol En- regulation of gluconeogenesis and lipid metabolism in males. Sci Rep 2017;7:1661 docrinol 2008;22:10–22 36. Onuma H, Vander Kooi BT, Boustead JN, Oeser JK, O’Brien RM. Correlation 16. Schmoll D, Walker KS, Alessi DR, et al. Regulation of glucose-6-phosphatase between FOXO1a (FKHR) and FOXO3a (FKHRL1) binding and the inhibition of basal gene expression by protein kinase Balpha and the forkhead transcription factor glucose-6-phosphatase catalytic subunit gene transcription by insulin. Mol En- FKHR. Evidence for insulin response unit-dependent and -independent effects of docrinol 2006;20:2831–2847 insulin on promoter activity. J Biol Chem 2000;275:36324–36333 37. Matsumoto M, Pocai A, Rossetti L, Depinho RA, Accili D. Impaired regulation 17. Ayala JE, Streeper RS, Desgrosellier JS, et al. Conservation of an insulin of hepatic glucose production in mice lacking the forkhead transcription factor response unit between mouse and human glucose-6-phosphatase catalytic Foxo1 in liver. Cell Metab 2007;6:208–216 subunit gene promoters: transcription factor FKHR binds the insulin response 38. Chen H, Hu B, Gacad MA, Adams JS. Cloning and expression of a novel sequence. Diabetes 1999;48:1885–1889 dominant-negative-acting estrogen response element-binding protein in the 18. Zhao X, Gan L, Pan H, et al. Multiple elements regulate nuclear/cytoplasmic heterogeneous nuclear ribonucleoprotein family. J Biol Chem 1998;273:31352– shuttling of FOXO1: characterization of phosphorylation- and 14-3-3-dependent 31357 and -independent mechanisms. Biochem J 2004;378:839–849 39. Yasrebi A, Rivera JA, Krumm EA, Yang JA, Roepke TA. Activation of estrogen 19. Guo S. Insulin signaling, resistance, and the metabolic syndrome: insights response element-independent ERa signaling protects female mice from diet- from mouse models into disease mechanisms. J Endocrinol 2014;220:T1–T23 induced obesity. Endocrinology 2017;158:319–334 304 Estrogen, Foxo1, and Glucose Homeostasis Diabetes Volume 68, February 2019

40. Schuur ER, Loktev AV, Sharma M, Sun Z, Roth RA, Weigel RJ. Ligand- 50. MacAulay K, Doble BW, Patel S, et al. Glycogen synthase kinase 3alpha- dependent interaction of estrogen receptor-alpha with members of the forkhead specific regulation of murine hepatic glycogen metabolism. Cell Metab 2007;6: transcription factor family. J Biol Chem 2001;276:33554–33560 329–337 41. Sisci D, Morelli C, Cascio S, et al. The estrogen receptor alpha:insulin re- 51. Murakami T, Shimomura Y, Fujitsuka N, Sokabe M, Okamura K, Sakamoto S. ceptor substrate 1 complex in breast cancer: structure-function relationships. Ann Enlargement glycogen store in rat liver and muscle by fructose-diet intake and Oncol 2007;18(Suppl. 6):vi81–vi85 exercise training. J Appl Physiol (1985) 1997;82:772–775 42. Morelli C, Garofalo C, Bartucci M, Surmacz E. Estrogen receptor-alpha 52. Lightfoot JT. Sex hormones’ regulation of rodent physical activity: a review. regulates the degradation of insulin receptor substrates 1 and 2 in breast cancer Int J Biol Sci 2008;4:126–132 cells. Oncogene 2003;22:4007–4016 53. Xu Y, Nedungadi TP, Zhu L, et al. Distinct hypothalamic neurons mediate 43. Sun M, Paciga JE, Feldman RI, et al. Phosphatidylinositol-3-OH Kinase (PI3K)/ estrogenic effects on energy homeostasis and reproduction. Cell Metab 2011;14: AKT2, activated in breast cancer, regulates and is induced by estrogen receptor 453–465 alpha (ERalpha) via interaction between ERalpha and PI3K. Cancer Res 2001;61: 54. Sadagurski M, Leshan RL, Patterson C, et al. IRS2 signaling in LepR-b 5985–5991 neurons suppresses FoxO1 to control energy balance independently of leptin 44. Simoncini T, Hafezi-Moghadam A, Brazil DP, Ley K, Chin WW, Liao JK. action. Cell Metab 2012;15:703–712 Interaction of oestrogen receptor with the regulatory subunit of phosphatidyli- 55. Matic M, Bryzgalova G, Gao H, et al. Estrogen signalling and the metabolic nositol-3-OH kinase. Nature 2000;407:538–541 syndrome: targeting the hepatic estrogen receptor alpha action. PLoS One 2013;8: 45. Zhu L, Xu P, Cao X, et al. The ERa-PI3K cascade in proopiomelanocortin e57458 progenitor neurons regulates feeding and glucose balance in female mice. En- 56. Simpson ER. Sources of estrogen and their importance. J Steroid Biochem docrinology 2015;156:4474–4491 Mol Biol 2003;86:225–230 46. Kumagai S, Holmäng A, Björntorp P. The effects of oestrogen and progesterone 57. Jones ME, Thorburn AW, Britt KL, et al. Aromatase-deficient (ArKO) mice on insulin sensitivity in female rats. Acta Physiol Scand 1993;149:91–97 have a phenotype of increased adiposity. Proc Natl Acad Sci U S A 2000;97: 47. Ahmed-Sorour H, Bailey CJ. Role of ovarian hormones in the long-term 12735–12740 control of glucose homeostasis, glycogen formation and gluconeogenesis. Ann 58. Jones ME, Boon WC, Proietto J, Simpson ER. Of mice and men: the evolv- Nutr Metab 1981;25:208–212 ing phenotype of aromatase deficiency. Trends Endocrinol Metab 2006;17:55– 48. Bitman J, Cecil HC, Mench ML, Wrenn TR. Kinetics of in vivo glycogen 64 synthesis in the estrogen-stimulated rat uterus. Endocrinology 1965;76:63–69 59. Haeusler RA, Pratt-Hyatt M, Welch CL, Klaassen CD, Accili D. Impaired 49. Roach PJ, Depaoli-Roach AA, Hurley TD, Tagliabracci VS. Glycogen and its generation of 12-hydroxylated bile acids links hepatic insulin signaling with metabolism: some new developments and old themes. Biochem J 2012;441:763–787 dyslipidemia. Cell Metab 2012;15:65–74